Approximately 60,000 new cases of Parkinson's disease are diagnosed each year in the US. Parkinson's disease is second in prevalence only to Alzheimer's disease among neurodegenerative disorders, and causes untold anguish to patients and their families, and costs patients and society billions of dollars per year. To facilitate better understanding of Parkinson's disease, the Michael J. Fox Foundation has sponsored The Parkinson Progression Marker Initiative. Toward this end, study evaluates hundreds of normal and Parkinson's subjects, and collects biological samples and MR images, and makes these data freely available to researchers. We propose to apply sophisticated methods for analyzing complex, large data sets to these data, to help researchers understand the underlying changes in genes and brain structure in patients in Parkinson's disease. We also propose to apply robust methods for automatically detecting subtypes among people with Parkinson's disease, based on brain connectivity as revealed by MRI, or by genetic differences. We expect that these differences will indicate differing prognosis, and perhaps point toward promising treatments.